[GP125] All that you wanted to know on Aprilia RSA 125, and more, by Mr Jan Thiel and Mr Frits Overmars (PART 3) (Locked)

hi

Frits did you ever posted somewhere The leaning tower of Pisa theory in German?

thanks Manuel

Got you that cleared it up nicely. Thanks for that.
Is there a disadvantage of holding it against the back wall to hard other than the fact its reducing area through the A duct due to the higher angle.
I would think you would want the less angle needed to get the job done correctly.This is where your software comes in as it makes a mathematical calculation
Where does one get a copy of your software
I normally try to aim the c port at the start of the combustion chamber closest to the C- port
my theory was it would help clean out the combustion chamber. I guess in this instance it would also be less likely to push the top of the coloum over or to try and separate it from that back wall.

Manuel Rainer a écrit:

Frits did you ever posted somewhere The leaning tower of Pisa theory in German?

Nein; zuviel zu tun und zu wenig Zeit.

hi frits i wonder if you can offer some advice for junior tuners like myself. i work on various kinds of 2t engines and im trying to get a better understanding of where a good starting point is regarding the timing and width of the windows that can relate to many cylinders. most of what i deal with is pretty conservative on the timing and width  from the manufacturer. rather than raise and widen the window gradually in every conceivable combination trying to find what is best , is there a good number that will be a good starting point that can save some time without going over board right away ?  for instance if i gave you a engine that you werent familiar with and said make it run good is there a particular timing/width you would immediatly use as the starting test point because you would know the factory timing is no good for best power ?  i thought once you had mentioned 190/130 with 70% width on the exh but i dont recall anything about the transfers width, so would this be a immediate good starting point ? thnx for any help.

Funny enough the cylinder I am porting at the moment it has the a ports flat and angled as per or very close to the drawing you have up there.
The b duct aims in very hard to the c duct and its angles up
the axial angles are back to front in comparison to your thoughts. Thats a good thing it means I can change it around and hopefully get a bit more out of it. From my understanding of the paper you wrote the setup I have would be tilting the column towards the exhaust port and out. This is of course far from ideal.I plan on bring the B duct and A duct further out towards the centre of the cylinder radially. I would be very interested in your thoughts on this
thanks again

[Vous devez être inscrit et connecté pour voir cette image]

brokedown a écrit:

hi frits i wonder if you can offer some advice for junior tuners like myself. i work on various kinds of 2t engines and im trying to get a better understanding of where a good starting point is regarding the timing and width of the windows that can relate to many cylinders. most of what i deal with is pretty conservative on the timing and width  from the manufacturer. rather than raise and widen the window gradually in every conceivable combination trying to find what is best , is there a good number that will be a good starting point that can save some time without going over board right away ?... i thought once you had mentioned 190/130 with 70% width on the exh but i dont recall anything about the transfers width, so would this be a immediate good starting point ?

If your main goal is to build fast engines, you can use the 190/130 timing with 70% main exhaust port width, make sure to add auxiliary exhaust ports, and copy the scavenging layout I posted above.
If you want to understand two-strokes, the step by step method will teach you much more before you finally end up with the same layout I just mentioned.

wax a écrit:

From my understanding of the paper you wrote the setup I have would be tilting the column towards the exhaust port and out.

You understood correctly, Wax. Luckily the B-ports in your setup have rather small effective areas; otherwise they would cause a massive loss of fresh mixture right into the exhaust port. But small effective areas also mean that only at low revs the cylinder will be adequately filled.

  May be somebody here could help me to solve my problem.

I put it there: [Vous devez être inscrit et connecté pour voir ce lien]

I would like to add a muffler to my 70's "racer"

may be it was not the right place, if so, could an admin move this post in a better place ?

end of   

Hello All:
Frits the drawing you iincluded in the Transfer Theory sheets you had axial angles of 25° for the a-ports, 25° for the b-ports, and 45° for the c-port with a radial scavenging direction resultant of 101.045°. You said anything over 90° the column was headed toward the exhaust port. In the drawing above you changed the angle of the b-port to 10° to better this resultant. You also changed the angle of the c-port from 45° to 50°. But I also see you narrowed the c-port. Was this to give some more width to the b-port that was lost because of the change to the flanks to the a-ports? I also see your new radial scavenging direction resultant is 100.622°. Why is this high if you changed the angle of the b-port? Or is it because of the increased angle of the c-port? Is this resultant also over 90° for the RSA?

You are right, Jeff. The scavenging picture that I used in the Pisa-story was from a rather peculiar engine, a 6,5 cc methanol-burning glowplug-ignition engine. At the time I had not yet prepared the more general scavenging picture that I posted here yesterday, so I used what was at hand.
But I also seem to remember that I already answered several of your questions at the bottom of page [Vous devez être inscrit et connecté pour voir ce lien]
The reason for narrowing the C-port was indeed to make more cylinder wall available for the B-ports.
The resultant for the RSA is also over 90°; it means that the scavenging column is leaning over toward the exhaust side, but only slightly so; it is still far from aiming at the exhaust ports.

Thanks Frits. you described the engine well. This is an engine that does not like to rev. I can change the angles a fair bit as it has very low transfer port timing and I need to raise it. In that I will fix the bad angles in the axial direction.
I will also try and move the b ducts a little to flow more into the cylinder. As always your insight has been very helpful.
I love these two stroke engines and learning even more about them is always exciting

Hi Frits,
As there is so very little time available at Peak Torque RPM (12500) on the RSA125, how much of the total volume of the transfers ports are actually moved into the cylinder in one cycle?
Regards
Allan

wax a écrit:

Funny enough the cylinder I am porting at the moment it has the a ports flat and angled as per or very close to the drawing you have up there.
The b duct aims in very hard to the c duct and its angles up
the axial angles are back to front in comparison to your thoughts. Thats a good thing it means I can change it around and hopefully get a bit more out of it. From my understanding of the paper you wrote the setup I have would be tilting the column towards the exhaust port and out. This is of course far from ideal.I plan on bring the B duct and A duct further out towards the centre of the cylinder radially. I would be very interested in your thoughts on this
thanks again

thats how one of mine was. 15* on the A and 25* on the B. after grinding and positioning the cylinder with extra gaskets the best i could do was 20* on the A and 12* on the B. also added epoxy so the B wasnt going hard into the C.  still i think i made errors on some the radial angles and timings so its back out to the garage to start over   

Is anyone willing to upload some picture of a RSA cylinder ?
And if so would you ?
And if it's not to much to ask would you show in the pictures the ports entry to the bore please

Thanks

RAW a écrit:

Is anyone willing to upload some picture of a RSA cylinder ?
And if so would you ?
And if it's not to much to ask would you show in the pictures the ports entry to the bore please

Thanks

There is heaps of pics on these threads you really need to have a look around.

brokedown a écrit:

wax a écrit:

Funny enough the cylinder I am porting at the moment it has the a ports flat and angled as per or very close to the drawing you have up there.
The b duct aims in very hard to the c duct and its angles up
the axial angles are back to front in comparison to your thoughts. Thats a good thing it means I can change it around and hopefully get a bit more out of it. From my understanding of the paper you wrote the setup I have would be tilting the column towards the exhaust port and out. This is of course far from ideal.I plan on bring the B duct and A duct further out towards the centre of the cylinder radially. I would be very interested in your thoughts on this
thanks again

thats how one of mine was. 15* on the A and 25* on the B. after grinding and positioning the cylinder with extra gaskets the best i could do was 20* on the A and 12* on the B. also added epoxy so the B wasnt going hard into the C.  still i think i made errors on some the radial angles and timings so its back out to the garage to start over   

Its a work in progress huh.
Im kinda hoping to to have that same thing, time will tell

wax a écrit:

RAW a écrit:

Is anyone willing to upload some picture of a RSA cylinder ?
And if so would you ?
And if it's not to much to ask would you show in the pictures the ports entry to the bore please

Thanks

There is heaps of pics on these threads you really need to have a look around.

Hi Wax, yes heaps of pics but are they RSA ? I'm not so sure

Jan Thiel a écrit:

I think al the fresh charge going into the cylinder is that what is already contained in the transfer ducts!
Of course also including the crankcase ducts.
When you think of it: would there be time for the charge under the piston, or between the crankwebs
to reach the transfer ports?
And also there is a too large distance!

Hello Jan, and hello Frits also.
This post of a few weeks ago has got me really intrigued, and I think it would be useful to elaborate on this further. I have a number of questions for you both:

1)Gordon Blair said he believed peak transfer flow was at, or just after BDC. The work he did at QUB was quite rigorous for its time. Do you believe his assertion to be correct, and how influential was Blair's work on your own understanding and methods? I suppose this question is especially for you Frits since you are also a master of software simulations.

2) If the transfer charge constitutes mainly what is already retained within the duct column, we should be able to exploit this phenomena further, maybe with elaborate piston design aimed more at 'pumping' than they way we currently think about it, and concentrating not only on transfer port exit Time x Area & Orientation, but port entry Time x Area & Orientation. This could be an opportunity to break new ground?

3) Given what you have both discovered about the shape of the A + B port bridge, and the need for a radius at port entry, why is the piston skirt still essentially a knife-edge, when it could be say 5mm thick at the transfer point, and relieved of its subsequent mass somewhere else somehow?

4) Frits, you say with a long connecting rod and rotary inlet, Helmholtz derived depressions can essentially pull fresh charge right through the cylinder from carburetor to exhaust chamber. Given this 'grand visualisation', is it still fair to discount transfer charge's interaction with the piston skirt as of low consequence to attaining ultimate performance?

I hope my summaries of your respective teachings has been accurate enough to make my questions sensible  
These may also answer GtG001 's Volumetric Efficiency question.
Again, I really appreciate all you are doing for the community here!

nick gill a écrit:

1)Gordon Blair said he believed peak transfer flow was at, or just after BDC. The work he did at QUB was quite rigorous for its time. Do you believe his assertion to be correct, and how influential was Blair's work on your own understanding and methods? I suppose this question is especially for you Frits since you are also a master of software simulations.

In order to answer this question, we must first define transfer flow.
Transfer flow will start with an initial velocity of zero. Flow velocity will increase as long as the pressure at the entry side of a duct is higher than the pressure at the exit side.
When entry and exit pressures are equal, the flow velocity has reached its maximum value. Subsequently entry pressure will drop even further and exit pressure will continue to rise, which will slow the flow velocity down.
Ideally the flow will come to a halt just when the transfer ports are being closed. Assuming a symmetrical situation, flow velocity will thus be at a maximum halfway between the transfers' opening and closing points, at BDC; one point for Blair.
But we are not really interested in flow velocity; we want mass transfer, which is flow velocity * mixture density * flow cross area. This area is also maximal at BDC: two points for Blair.
Blair's work certainly had an influence om my understanding, although not so much on my methods. Blair developed a very elegant method for transient particle flow calculation. But Jan Thiel will argue that Blair never built a fast engine in his entire life; he was first and foremost a theoretician.

Citation :

2) If the transfer charge constitutes mainly what is already retained within the duct column, we should be able to exploit this phenomena further, maybe with elaborate piston design aimed more at 'pumping' than they way we currently think about it, and concentrating not only on transfer port exit Time x Area & Orientation, but port entry Time x Area & Orientation. This could be an opportunity to break new ground?

Transfer duct entry conditions are important, but because of the larger entry cross area and consequently lower flow velocity they are not nearly as important as the exit conditions.
Furthermore, the transfer ports are only open about 1/3 of the time during each crankshaft revolution, and most of this time they are only partially open; the entry areas are completely open all the time.
New ground? Concentrating on pumping is old-school; in the early fifties of the previous century people began to realize that a competition two-stroke engine should be regarded as an oscillating-flow device, not as a pump.

Citation :

3) Given what you have both discovered about the shape of the A + B port bridge, and the need for a radius at port entry, why is the piston skirt still essentially a knife-edge, when it could be say 5mm thick at the transfer point, and relieved of its subsequent mass somewhere else somehow?

If you could indicate where a modern piston could be relieved of its mass, you would make us very happy.

Citation :

4) Frits, you say with a long connecting rod and rotary inlet, Helmholtz derived depressions can essentially pull fresh charge right through the cylinder from carburetor to exhaust chamber. Given this 'grand visualisation', is it still fair to discount transfer charge's interaction with the piston skirt as of low consequence to attaining ultimate performance?

I did not assert that a long con rod and a rotary inlet are necessary for this phenomenon; it also works with short rods, with reed valves and even without any valving at all, provided the pressure fluctuations are in sync with the port timing, for which the piston is indispensable.

hey frits what your thoughts about slow vs fast burning fuel . like methanol vs gas. does it change the amount of blowdown needed ?

Hi Frits and Jan,
I have enjoyed the discussions over the life of this article and been challenged with the insights of yourself, Jan and other contributors – thank you very much to all.      

The two quotes below seem to also beg the question I asked a few days ago: does the mass transfer rate of the Transfer ports = 100% at Peak torque - or are the individual Transfer ports oscillating at different frequencies to each other based on their volume/length/RPM range and delivering a varying percentage of their mixture each cycle.
If they are, then the model for scavenging is not as balanced as it seems and is changing throughout the RPM range.

Jan Thiel a écrit:

I think al the fresh charge going into the cylinder is that what is already contained in the transfer ducts!
Of course also including the crankcase ducts.
When you think of it: would there be time for the charge under the piston, or between the crankwebs
to reach the transfer ports?
And also there is a too large distance!

Citation :
Furthermore, the transfer ports are only open about 1/3 of the time during each crankshaft revolution, and most of this time they are only partially open;

a competition two-stroke engine should be regarded as an oscillating-flow device

Another quote from Fritz(though not verbatim) ; a drop of fuel takes approximately 4 cycles to appear in the combustion chamber.

GtG001 a écrit:

does the mass transfer rate of the Transfer  ports = 100% at Peak torque - or are the individual Transfer ports oscillating at different frequencies to each other based on their volume/length/RPM range and delivering a varying percentage of their mixture each cycle.
If they are, then the model for scavenging is not as balanced as it seems and is changing throughout the RPM range.

Allan, the amount of mixture mass, passing through the cylinder at peak torque, depends on a lot of factors. The same goes for the amount of mixture mass, contained in the transfer ducts. Both amounts may or may not be equal; I can't say, because I never programmed my sim to display this ratio on the screen.

The pressure differential between crankcase side and cylinder side may be assumed as uniform for the various transfer ducts, but individual ducts may differ in their respective volumes and lengths, so the acceleration of flow may differ. This difference ratio does however not vary with rpm, so the scavenging balance does not vary (assuming all transfer ports are of the same height and blowdown into them does not occur).
The ports have no different oscillating frequencies; the complete engine has a compound frequency. I tried to explain that in a little story called Helmholtz Blues. You can find it here: [Vous devez être inscrit et connecté pour voir ce lien]

Citation :

Jan Thiel a écrit:
I think al the fresh charge going into the cylinder is that what is already contained in the transfer ducts!
Of course also including the crankcase ducts. When you think of it: would there be time for the charge under the piston, or between the crankwebs to reach the transfer ports?

The charge under the piston does not move nearly as much as the charge in the crankcase and the transfer ducts.
That is a pity because if it did, it would do a better job of cooling the underside of the piston.
We once used a stationary stuffer, mounted above the crankshaft, that reached up into the piston. Despite reducing the crankcase volume, it had a positive effect on power, maybe because it caused more movement of this under-piston mixture.

Hi Frits,
Thank you for your reply.  
I did read your article on the Helmholtz blues originally as I do with all you write here and that is one of the reasons I asked the question. As the transfers are but a port between two resonators which vary in volume with each cycle and they are also connected to the outside world by other varying resonators, there seems that nothing can be constant within the transfer ports both in frequency and mass flow. As we know the transfers port open area is changing with each cycle so wouldn’t the frequency change also throughout the duration of the transfer port?  
Please excuse my ignorance for pursuing this on line of thought.   

I referred to my Helmholtz Blues because I wanted to make clear that a duct alone cannot have a frequency; the volumes at both sides of the duct also play a role. Such a volume can be very large (the outside world) or very small (a closed end) but there is always more than just that duct.
In an engine the volumes are coupled to other volumes via other ducts; the coupling areas are constantly changing; some of the volumes themselves are changing; the local temperature and thus the speed of sound vary, and so on. That is why I talk about a compound frequency for the whole engine.

A frequency is the number of cycles of varying pressures within one second. Within such a cycle the speed of this variation can fluctuate; for example it can resemble a sine curve, but it might also resemble a different wave shape. Maybe that is what you meant?

I think we should not dive too deep into this subject here, Allan, or we might loose all our readers. After all, this is a motorcycle forum, not a physics lecture.

Thanks Frits, very informative, shame you don't want to go further down this road, it's the 1st time I've been able to visualise this concept and the thought of being taught more directly associated with the 2stroke engine would be a great read